Locating desirable resources and information from a large-scale grid is challenging due to the considerable diversity, large number, dynamic behavior, and geographical distribution of the resources. In this paper, we propose an efficient discovery framework which organizes a grid network by a semantically linked overlay representing the semantic relationships between grid participants. Specifically, we use a semantics-aware topology construction method to group similar nodes to form a semantic

more »

... mallworld. With the small-world topology constructed, resource-discovery queries will be propagated only between semantically related nodes, which greatly improves the efficiency and accuracy of resource discovery in grids. Moreover, we propose a novel algorithm for efficient resource information integration and searching over the semantic small-worlds. Our experiments with simulations substantiate that this framework significantly improve the search expressiveness, efficiency, scalability, and precision.